African Journal of Basic & Applied Sciences 9 (1): 07-18, 2017ISSN 2079-2034© IDOSI Publications, 2017DOI: 10.5829/idosi.ajbas.2017.07.18

Corresponding Author: Mehreen Shah, SZABIST, Dubai, United Arab Emirates7

Service Quality, Customer Attitude and Subscription Behavior: A Structural EquationModeling Approach to Determine Mediating Relationship

Mehreen Shah, Nida Tanveer and Zubair Azam

SZABIST, Dubai, United Arab Emirates

Abstract: The research study aims to investigate the nature, strength and direction of association betweenservice quality and customer subscription behavior with attitude’s mediation role. To achieve stated objective,causal research was designed where quantitative data was collected through self-administered questionnairesand 416 observations were collected from customers surveyed using convenience sampling. Instrument wasadapted. Firstly, Regression Analysis with PROCESS Add-in was conducted to quantify the associationalrelationships. 2 -order factor-based structural regression Model has been developed and tested for the study.nd

Data analysis proves partial mediation. Service quality was having 92% positive direct influence on Customersubscription behavior but reduced to 39% because of mediating effect of attitude. This study offers newempirical support for the proposition that attitude towards service provider play mediating role betweenperceived service quality and subscription behavior. As such, findings from this study aspires managers todevelop attitudinal strategies in order to increase customer retention and acquisition in telecommunicationservices. The findings can be generalized by expanding the industrial dynamics in sampling frame. The researchstudy significantly contributes to existing body of service management literature because of its inherentempirical nature. The findings can be applied in variety of Subscription-base industries including magazines,websites, online social communities, newspapers, cable T.V, event management and financial services.

Key words: Service Quality Consumer Attitude Subscription Behavior Telecommunications Industry Structural equation modeling

INTRODUCTION impact on the buyer subscription behavior. Service

Subscriptions are not a new form of retailing but they perceived quality and expectations [3]. are in practice from a long time [1]. Many industries were Several authors [4, 5] have carried out researchbenefited from subscription mode of retailing. initiatives to find impact of service quality on variousSubscription Behavior has been defined as purchasing of dimensions of consumer behavior in various industries.products and services as a regular or for some specific However, each study exclusively focused on direct effectstime period [2]. Goods and services sold via subscriptions of five dimensions of service quality proposed ininclude printing industry, social networking sites, SERVQUAL model [6]. Service quality’s indirect effect onpodcasts, health club membership, telecommunication customer behavior has not been tested and generalized onservices and many household items. It’s common in broader level especially with customer’s attitude towardsvarious consumer and client industries to provide Telecom service provider as a mediator. Furthermore, it’sincentives to new members in the form of discounts and evident from the phase of desk research that the buyertrial to attract the customers in order to increase subscription behavior in telecommunication services hassubscription rate. not been explored in the U.A.E.

A notable demand for subscription rate is being feltas markets are getting more and more fragmented and Research Objectives: The present research intends tomost of commodities are now considered homogenous in analyze the indirect effect of service quality onnature, hence the service quality may have a notable subscription behavior with attitude towards telecom

Quality is the degree or direction of discrepancy between

African J. Basic & Appl. Sci., 9 (1): 07-18, 2017

8

services as a mediating variable. Telecommunication have to be very careful in controlling the entiremarket of UAE is perceived as one of the largest target environment as mistake in service cannot be unfold evenmarket in emerging economies. An extensive amount of if corrected later, as it leaves an impact on customers mindresearch has been and is being conducted all over the [11].world to quantitatively determine the service quality’scontribution to buyer subscription behavior, however, it Service Quality: Defining the notion of service qualityis important to disclose that sufficient research studied initially starts with a customer orientation, as quality ishave not been conducted to explore consumer dynamics perceived by the customers and if meets or exceeds hisin UAE which makes it a very viable area of research. perceptions further leads to profitability [12]. Service

Literature Review discrepancy between perceived quality and theMarketing Concept and Subscription Behavior: expectation [6]. It’s obvious that if the perceived qualityConsumer behavior research has witnessed a tremendous is same as it was expected, the service quality [13] will begrowth in the last 65 years since the inception of 'The positive resulting in satisfied customers [14, 15]. However,marketing concept' [7]. The Marketing concept asserts if the perceived quality is not as per the expectations,importance of Customer as the focal stakeholder and negative attitude of the customers towards the service willcentre of every marketing decision in any organization result in customer loss [14, 15]. Service quality is mostaround the globe [7, 8]. These customers have been commonly defined as difference between perceived andcategorized as business clients and individual consumers. expected service [16]. Past studies have proved theBusiness customers are the clients who enjoy an umbrella positive relationship between service quality andof contractual arrangement while individual consumers are satisfaction [17, 18].the ones who purchase some goods and services for Customer Expectations serve as benchmark andpersonal or home usage [7]. Many of individual, home or foundations to access whether product or services isbusiness services are acquired by these customers quality oriented or not [19]. If performance of product andthrough paid subscriptions. services meets customer expectations, it is considered

Quality has been considered a tool to drive customer quality oriented and if it falls short of expectations thansatisfaction in both products oriented and service customer does not perceive it quality oriented [20].oriented firms [9]. A services provider, while inquiringabout the potential problem they might face in servicesprovision process, it can focus on three tries/levels inorder to comprehend the ‘Service Game’ and can devisestrategies accordingly to and up as a service winner andchampion [9, 10].

The three tier view is the bottom to top approach interm of importance of each tier. The very import tier of anyservice organization is the CUSTOMER TIER. Customers Fig. 1: Three tiered view of service Organizationare considered as the foundation of service organizationand customers are not external to the organization they Measuring Service Quality: Service quality deals withare the part of organization. The customer tier is analyzed interaction between customer and service provider. It isin term of expectation for quality and need for security, very important to determine whether the perceived qualityesteem and justice. is positive or negative. There are two major approaches

The second important tier is the BOUNDARY TIER being used by the service providers to estimate thewhich is conceptualized as everything with which Service quality which are Passive Approach and Activecustomer came in contact during service encounter approach [13]. In Passive approach the comments card areincluding people, equipment and the overall ambience. available at the counter, customers can pick up the

The third and last important tier is COORDINATION comment card and fill it out if they wish to, whereas, theTIER. This tier is responsible for managing all the entire active approach is the formal survey and interview takenservice climate and departments like marketing, from the customers. The former approach is considered tooperations, finance, human resource etc. The management be more accurate and reliable [21].

quality is also defined as the degree or direction of

African J. Basic & Appl. Sci., 9 (1): 07-18, 2017

9

Table 1: Application of Service Dimensions in Telecom SectorService Dimension Conceptualization and Application1. Tangibility The appearance of physical facilities, equipment, personnel and communication material of an organization e.g.,

Number of service centers, location of centers, counters in each service centre, design and layout of centers and Aesthetics ofCenters along with employees.

2. Reliability Ability of employees to perform the promised service dependably and accurately is the reliability e.g., number of complaintsreceived versus number of complaints resolved versus number of complaints pending.

3. Responsiveness Ability of Employees to provide timely and prompt responses e.g., waiting time for customers in physical and online settings.4. Assurance Ability of employees to assure customers about solvability of their issues and complaints e.g., knowledge about operating

procedures and processes. 5. Empathy Attitudes, communication and caring behavior of employees when they are having interactions with customers on any medium of access.

A perceived service is, thus, the outcome of how the study is conducted on service quality, the gap is to becustomers judge different dimensions of services quality. examined between expectation and perception of theThe Gaps Model [21, 22] continues to dominate service received by the service provider [31]. amongst the service analysis and improvementframeworks. It determines where does the problems lies, Servqual: Parasuraman (1985/1988) designed thewhat are the root causes and potential consequences and SERVQUAL model. This model helps in betterthen suggest how to resolve them in order to increase understanding the difference between expectations andcustomer satisfaction. perception of the customer [13]. The model contains 22

There are different models to quantify the service item questions which can be used in any service sectorquality of products and services. The two essential with some adjustments and changes. This model coversapproaches to conceptualizing and measuring service 5 dimensions which are tangibility, reliability,quality are SERVQUAL model and SERVPERF model [23]. responsiveness, assurance and empathy. The dataThe SERVPERF model was proposed by Cronin & Taylor collected through this model assist the service provider to(1992/1994) and it distinguishes perceived quality from evaluate the difference between the expected serviceexpected quality, whereas, the SERVQUAL model which quality and perceived service quality. There are 5 Gaps inwas later known as rater model is a quality management the SERVQUAL; these gaps are between the expectationframework [3]. and actual outcome of quality

Parasuraman et al., (1988) proposed SERVQUALmethodology consisting of five dimensions (reliability, GAP1: Gap between consumer expectations andassurance, tangibles, empathy and responsibility). Which management perceptions.have been a core service quality analysis toolkit and have GAP2: Gap between management perception and SQbeen used, adopted and revised by many later scholars specification.for their particular domains [24, 29]. The Table 1 presents GAP3: Gap between SQ specification and service delivery.conceptualization of 5 dimensions of services quality with GAP4: Gap between service delivery and externalrespect to telecommunication industry. Managerial communication.choices about any of 5 services dimensions can positively GAP5: Gap between perceived service and expectedor negatively affect customer attitude towards service service.provider.

This approach contains five elements i.e. reliability, Attitude: Attitude has been studied by socialassurance, tangibles, empathy and responsiveness. A psychologist for last many years decade’s. Marketersquestionnaire is used to measure both the services have used the concept of attitude to study the nowexpectation and perception [3, 13]. When service is consumer respond to the products and servicesperceived as positive, it will attract the customers to offered[25-40].repeat the buying behavior, as buying behavior increase Responses are classified in three types as actions,the loyalty factor in the customers would arise [29, 30]. feelings and thoughts according to Tripartite Model [32].

In modern growing economy and increasing Later, Ajzen & Fishbein, (1980) agreed with the Triandiscompetition, SQ is considered very essential for the (1980) argument that attitude represents person favorablesuccess and survival of the organization [22]. Whenever and unfavorable feelings.

African J. Basic & Appl. Sci., 9 (1): 07-18, 2017

10

Attitude is person overall evolution of an objective (1965) suggests winning and then retaining a customer asit can be positive or negative. Evaluation is done by sole purpose of business and employing marketingeffective and cognitive system. Effective system consist strategies are, primarily, employed to win the customersof emotions, feelings, moods while Cognitive system share of wallet and heart. The success of marketingrefers to of consumer overall evolution based on initiatives is determined through their ability of getting toknowledge, meanings and beliefs about object [15]. know, entertain and influence customers in order toAttitude is based on the knowledge about market offering convert them from a prospect to a loyal customer [13].[21, 34, 35]. Furthermore, a model of service consumption was

Attitude has been always been a primary research proposed by Lovelock and Wirtz (2007/2010) which hadarea for the all types of psycho logistics, however, it had pre-purchase stage, service encounter stage andno formal definition till year 1935. Allport (1935) offers a post-purchase stages as three major stages a customerdefinition of attitude after understanding and reviewing goes through to purchase any high-involvement or16 definition of attitude which have been developed and low-involvement service [41-56].explain by previous researchers. Attitude, according to Pre-purchase stage refers to need recognition andAllport, (1935), is “A mental and neural state of readiness, information search for alternatives to alleviate that needorganized through experience, asserting a directive or [13, 23]. In the Second stage, potential solutions ordynamic influence upon the individual’s response to alternatives are evaluated and compared and thenall objects and situations with which it is related”. customer places order for the desired service. This stageSince then, researchers have evaluated, favored and is known as service encounter stage becauseextended this definition of attitude by applying it on customer has encounter with producer, environmentvarious fields of study e.g. marketing and organizational and other customers in the service centre and is thusbehavior. Eagly and Chaiken (1993, 1998) describe known as moment of truth [13, 23]. In the third stage,attitudes as “a psychological tendency that is expressed post-consumption evaluation of the service is carried out.by evaluating a particular entity with some degree of favor Customer compares service’s actual performance withor disfavor”. Prominent scholars have described attitude desired performance and depending on the zone ofin variety of way, however, they have developed tolerance, develops positive or negative intentionsconsensus on concept of evaluation being central to towards future consumption of the service. This is thedefinition of attitude. Attitudes, thus, refers to a person’s stage in which satisfaction or dissatisfaction willoverall evaluation of objects, persons, organizations occur and the loyalty decision will be made by theevents which may result in positive and negative customer [13, 23].responses [37]. In the process of purchasing or subscribing to a

There have been many approaches to determine how mobile telephony service, the customer assesses thedo consumers evaluate the brands and what shape their mobile service based on his/her evaluation of expectedresponse. Marketers improve service quality by focusing benefits versus risks and whether product/serviceon reliability of their services offering. Customer expects attributes has fulfilled, exceeded or failed to satisfythe performance of service provider to be consistent and expected bottom line. If service quality satisfiesdependable [7, 13]. Marketers need to provided services customer's expectations, customer will be satisfied andon time and whenever needed by customers. Moreover, may return to purchase in future and if it exceeds hisemployees must be able to reduce customer dissonance expectations, he will surely be returning back [13, 14].whenever a problem a rises and they need to assure However, if the expectations are not fulfilled by servicecustomer about solvability of complaint in a caring way in performance, customer will churn [13].order to avoid or reduce customer chaos [38].

Customers encounter with service personnel, service MATERIALS AND METHODSitself, other customers waiting to be served in serviceoutlet affect customer attitude toward the company and Research design refers to the plan, strategies,his intention to subscribe & generate word of mouth techniques and tools used to conduct investigation ofincreases or decreases accordingly [39]. any research problem [40]. The research is causal in

Customer Subscription Behavior: Subscription behavior relationship between one independent, one dependanthas been defined as purchasing of products/services as variable and one mediating variables. Researcher'sa regular or for some specific time period [2]. Drucker dependence on primary data makes it primary research,

nature as it intends to explain the cause-and-effect

African J. Basic & Appl. Sci., 9 (1): 07-18, 2017

11

although this method of research requires lots of time and Considering the suggestions [43, 44, 45], a sample ofcost but it increases the quality of research being 300 observations was decided. However, a total of 416conducted. By using the primary data collection method, observations were collected. The sampling techniquethe researchers plan to have a tighter control over the employed was convenience sampling. Though it is a non-research [57]. probability sampling technique, but was considered

As, this research intends to inspect the effect of the suitable because of time and cost constraints.service quality on customer subscription behavior along The data was reviewed and incomplete orwith customer attitude’s mediation role in stimulating this inappropriate responses were eliminated after dataeffect, quantitative research techniques are needed for screening. Missing Values Analysis (MVA) was carriedboth data collection and data analysis. In order to collect out in SPSS and Expectation Maximization was used anddata, survey technique is considered reliable one and missing values were replaced with series mean.thus, questionnaire has been adapted from the existingbody of knowledge. Research Instrument: This research was a primary in

Hypotheses: As discussed above, the researchers intend instrument was adapted from [2, 6, 13, 25, 46] and ato investigate the nature, strength and direction of self-administered questionnaire consisting of 2 parts wasassociation between service quality, attitude and buyer developed. The first part contained 5 demographicsubscription behavior. In order to fulfill the stated questions while the inferential part was subdivided into 3objective, researchers developed following hypotheses. categories mainly items belonging to Service Quality,

H1: Service Quality positively impacts buyer subscription summated rating scale, which is a variation of Likert-scale,behavior. was used where 5 refer to Strongly Agree and 1 refers toH2: Service Quality positively impacts customer’s Strongly Disagree. The design of the questionnaire is asattitude towards Telecom Service Provider. follows. As discussed above, questionnaire starts withH3: Attitude positively impacts buyer subscription demographic questions where first question asked genderbehavior. of the respondent. The second question was ofH4: Service quality positively impacts buyer subscription employment status. The third question probed aboutbehavior mediated by Customer’s attitude towards service preferred service provider. The forth question was aboutprovider. preferred packages. The fifth question was of preferred

Sampling Design: Respondents are all those people or divided according to three variables; services quality,participants who are involved in research for the purpose attitude and subscription behavior. Service quality wasof extraction of information [41]. Respondents of this further divided into five factors on the bases ofresearch include employees who work in organizations SERVQUAL model proposed by [47, 48]. SERVQUAL isincluding DU telecommunication, KHDA and students the active approach to measure the SQ. This approach isfrom different universities. A total of 416 observations known to be more valid and standardized instrument. Thewere collected using self-administered questionnaires and use of this instrument is very common in the retailresults were deduced from them. industry i.e. banks, hotels, restaurants, hospitals,

Stevens (2009) suggests that a good rule of thumb in telecommunication, travel agencies and libraries [21, 48,a multiple regression analysis is 15 cases per predictor. 49]. Attitude had three items borrowed from ABC modelIt is further suggested that in associational analyses, of attitude formation. Subscription behavior identifies fiveresearcher can even go for five cases per variable but only items which have been adopted from [37, 46, 50].if the distribution of collected data is perfectly normal andthere are no outliners [42, 43]. So it is very important to Reliability and Validity of the Research Instrument:distinguish that the Bentler and Chou (1987) and Steven’s As discussed above, the survey instrument was adapted(2009) recommendations slot in at about five to fifteen from existing body of knowledge. However, it iscases for one variable. Furthermore, many researchers absolutely necessary to establish construct validity alongsuggest a minimum of 10 observations per item in the with reliability [43, 45]. For the purpose of establishingquestionnaire [44, 45]. reliability and validity, researchers have used Cronbach’s

nature as the data was collected first hand. The research

Attitude and Customer Subscription Behavior. A 5-point

type of service. The second part of questionnaire was

African J. Basic & Appl. Sci., 9 (1): 07-18, 2017

12

Table 2: Reliability and Validity analysis

Variables Cronbach's Alpha CR AVE MSV

Service Quality .928 0.875 0.792 0.537

Attitude 0.780 0.773 0.729 0.461

CSB 0.852 0.824 0.764 0.528

Alpha, composite reliability (CR), average varianceextracted (AVE) were used to assess convergent validity.AVE is considered an outdated measure of convergentvalidity. Malhotra and Dash (2011; p.702) comment intheir book as “AVE is a more conservative measure thanCR. On the basis of CR alone, the researcher mayconclude that the convergent validity of the construct isadequate, even though more than 50% of the variance isdue to error.” Cronbach’s Alpha, on the other hand,ensures the internal consistency of the questionsand responses. The values provided in Table 2substantiate that researchers have ensured overallaggregate reliability.

Data Collection Procedure: The data collection processfor this research was that surveys were physicallydistributed. The researchers distributed it along withpen given to respondents and with cover letterexplaining a brief background of research andensuring confidentiality. Respondents could fill thequestionnaire and immediately the researcher could viewresponses.

Statistical Tool Analysis: Researchers used descriptivestatistics to uncover perception patterns in Nominal andordinal variables. However, multiple regression analysiswas performed in order to find the cause-and-effectrelationship between SQ and customer SubscriptionBehavior with attitude as a mediating variable. To checkthe mediating effect of attitude, researchers’ used 3techniques in consideration which were Sobel’s Test,Bootstrapping and Structure equation modeling (SEM).Sobel [51] proposed an approach of multiplying tworegression coefficients obtained in Model 1 and Model 2.However, Sobel’s test is an outdated approach and isconsidered a flawed one because of its inability to analyzecomplex models. Bootstrapping is another technique tocheck the mediation models but researchers decided tonot use it because of its non-parametric nature.Researchers used Structure equation modeling (SEM) todo a path analysis of conceptual framework. Table 3provides a summary of all statistical techniques used toconduct the analysis.

RESULTS

Hypotheses Testing: Theory claimed causal relationshipbetween SQ and subscription behavior mediated bycustomer attitude towards service provider. In mediationanalysis, 2 models surface. Model 1 claims the directeffect of predictor (SQ) on outcome variable (CSB) whilemodel 2 determines indirect effect of SQ on CSB withmediating effect of attitude (M). In order to testhypothesis, researchers [43, 45, 52, 53] suggest the use ofparametric tests because of greater accuracy, authenticityand inferential power. However, parametric statisticsrequire certain assumptions to be checked in order toproceed further. Any violation of these assumptions doesnot allow researchers to move ahead and thus, non-parametric tests will be utilized. Table 3 provides asummary of assessment checks.

Assumptions of Parametric Statistics: As illustrated inTable 3, parametric statistical testing requires the data tohave 6 characteristics. Firstly, the scale used to collectdata shall have internal consistency and compositereliability. It has been tested using Cronbach Alpha, CR,AVE and MSV. The results (Table 2) revealed a reliableand valid measurement scale. Secondly, parametricstatistics require sample adequacy and was tested usingitem-based rule, KMO & Bartlett’s test and HOELTERtest. The results indicated data adequacy and sufficiency.Thirdly, data shall be normally distributed and it can betested using Skewness and Kurtosis as well asKolmogorove-Smirnov test. All of the values of skewnessand kurtosis were within the range of +1 to -1.

Hence, data was found normally distributed.Fourthly, parametric statistics require data to havelinearity and is usually tested using residual plots withFitline. Residual plot indicated a perfectly linear line andheteroskedasticity was absent.

A very important assumption of causal analysis isabsent of multicollinearity and is usually tested usingTolerance values, VIF values and condition Index. Resultsobtained in regression indicated no multicollinearity.Lastly, there shall be no correlation among independentvariables and it is tested through Durbin Watson test.When applied, value of 1.938 indicated no presence ofautocorrelation.

Mediation Analysis Using Process: When analysis wasexecuted using PROCESS Ad-on and results are shown inTable 4, Model 1 yielded significant relationship (p=0.00)

African J. Basic & Appl. Sci., 9 (1): 07-18, 2017

13

Table 3: Parametric Statistics Assumptions check

Analysis / Test Technique Finding Conclusion

1. Reliability Analysis Cronbach Alpha Above 0.59. Scale is Reliable and has no validity problem.CR Above 0.70AVE Above 0.50MSV Less than AVE

2. Sample Adequacy Item-Based Rule 30 * 10 = 300 < 416 Sample Size is sufficient and AdequateKMO & Bartlett’s Test 0.850

P = 0.00HOELTER (0.05 & 0.01) 244 & 256 < 416

3. Data Normality Skewness & Kurtosis.Histograms. Values are between -1 to +1. Data is normally distributed.

4. Multicollinearity Tolerance (>1-R ) Tolerance values are above 1-R . No multicollinearity.2 2

VIF VIF values are below 3.Condition Index Values are below 30.

5. Heteroskedasticity Residuals Plot with Fit Line Linear line No Heteroskedasticity6. Autocorrelation Durbin-Watson 1.938 No Autocorrelation7. Casual Analysis Regression using PROCESS ad-on When attitude was entered as mediator, R value reduced Partial Mediation2

but p-value remained significant.

Table 4: Regression Results Table 5: Item-based Factor Loadings

Model = 4Y = CSBX = SQM = AttitudeSample size416

Outcome: Attitude

Model 1 SummaryR R-sq F p0.6998 0.4898 359.01 0.0000

Model 1 Coeff T pConstant 0.3527 1.8483 0.0654SQ 0.9277 18.9478 0.0000

Outcome: CSBModel 2 SummaryR R-sq F p0.7582 0.5749 252.21 0.0000

Model 2 Coeff T pConstant -0.4353 -2.2468 0.0252Attitude 0.3932 7.5220 0.0000SQ 0.6826 9.8520 0.0000

Direct and Indirect effectsDirect effect of X on YEffect T P0.6826 9.8520 0.0000

Indirect effect of X on YMediating Variable Effect BootLLCI BootULCIAttitude 0.3647 0.2474 0.4904

Partially Standardized Indirect Effect of X on YMediating Variable Effect BootLLCI BootULCIAttitude 0.3978 0.2688 0.5252

Completely Standerdized Indirect effect of X on YMediating Variable Effect BootLLCI BootULCIAttitude 0.2488 0.1757 0.3263

Items Factor-loadings P-ValueT1 (Tangibility) 0.53 0.000T2 (Tangibility) 0.61 0.000T3 (Tangibility) 0.63 0.000RL1 (Reliability) 0.57 0.000RL2 (Reliability) 0.63 0.000RL3 (Reliability) 0.66 0.000RL4 (Reliability) 0.65 0.000RS1 (Responsiveness) 0.75 0.000RS2 (Responsiveness) 0.74 0.000RS3 (Responsiveness) 0.73 0.000RS4 (Responsiveness) 0.56 0.000RS5 (Responsiveness) 0.58 0.000AS1 (Assurance) 0.72 0.000AS2 (Assurance) 0.76 0.000AS3 (Assurance) 0.77 0.000AS4 (Assurance) 0.71 0.000AS5 (Assurance) 0.74 0.000EM1 (Empathy) 0.72 0.000EM2 (Empathy) 0.69 0.000EM3 (Empathy) 0.64 0.000EM4 (Empathy) 0.66 0.000EM5 (Empathy) 0.65 0.000A1 (Attitude) 0.77 0.000A2 (Attitude) 0.69 0.000A3 (Attitude) 0.76 0.000SB1 (Subscription Behaviour) 0.78 0.000SB2 (Subscription Behaviour) 0.75 0.000SB3 (Subscription Behaviour) 0.56 0.000SB4 (Subscription Behaviour) 0.82 0.000SB5 (Subscription Behaviour) 0.80 0.000

with Value of 0.048. SQ positively (t = 18.94) impacts2

customer attitude and causes a change of 92% ( = 0.92).Analysis of model 2 revealed partial mediation becauseindirect effect of SQ on CSB was reduced from 0.6826 to0.3647 while p-Value (p = 0.00) was still significant. Theresult was further substantiated by partially standardizedeffect (0.39) and completely standardized effect (0.24).

African J. Basic & Appl. Sci., 9 (1): 07-18, 2017

14

Structural Regression Model: Confirmatory factor represent strong association. The -Value for H1 is 0.790analysis (CFA) was performed to measure the data fitness representing a change of 79 % in CSB because of oneof latent variables with their constructs. Table 5 shows degree change in SQ. Therefore, hypothesis I claimingestimates along with P-Value of each construct and is positive impact of SQ on CSB is proved and accepted.reflecting strong factor loadings (p = 0.3) along with The regression weight between customer attitude andsignificant results (p = 0.05). CFA results strengthen the service quality is having a value of 0.832 revealing 83%belief of inter-item consistency and true representation as positive change being caused by any change in SQ andno item was excluded from the instrument. thus, proving hypothesis II as well.

Structural regression model has been found Moreover, hypothesis III claiming relationshipexhibiting goodness-of-Fit. Researchers [43, 44, 45] between CA and CSB has also been proved as p-value issuggest CMIN/DF ( 3.0), CFI ( 0.9), RMSEA ( 0.05), GFI below 0.05 with -Value of 0.708 revealing 70.8% positive( 0.9) and AGFI ( 0.8). The results (Table 6) clearly change being caused in CSB by CA.prove goodness-of Fit of hypothesized model as value of Lastly, hypothesis IV which claimed mediating effectall measures are greater than bottom-line values of customer attitude between SQ & CSB has been(CMIN/DF = 1.931, P = 0.00, CFI = 0.941, RMSEA = 0.047 accepted. Results show partial mediation as direct effectand AGFI = 0.87) of SQ on CSB has been reduced from 0.790 to 0.338 upon

All Four Hypotheses have been proved as p-Value addition of attitude as a mediator. However p-value (0.03)for all causal relationships are below 0.05 and hence, represents a significant partial mediating relationship.

Table 6: Goodness-of-Fit Statistics for Measurement ModelModel NPAR CMIN DF P CMIN/DFDefault Model 83 737.832 382 0.000 1.931Saturated model 465 .000 0Independence model 30 6426.781 435 .000 14.774Baseline ComparisonModel NFI Delta1 RFI rho1 IFI Delta 2 TLI rho2 CFIDefault model 0.885 0.869 0.941 0.932 0.941Saturated model 1.000 1.000 1.000Independence model 0.000 0.000 0.000 0.000 0.000RMSEAModel RMSEA LO 90 HI 90 PCLOSEDefault model 0.047 0.042 0.052 0.797Independence model 0.182 0.178 0.186 0.000

Table 7: Regression WeightsDV IV Estimate S.E. C.R. PCustomer Subscription Behaviour <------- Service Quality 1.748 0.224 7.819 ***Attitude <------- Service Quality 0.604 0.042 14.272 ***Customer Subscription Behaviour <------- Attitude 0.708 0.164 4.323 ***Customer Subscription Behaviour <------- Service Quality 0.317 0.107 2.975 0.003

Table 8: Standardized Regression WeightsDV IV EstimateCustomer Subscription Behaviour <------- Service Quality 0.790Attitude <------- Service Quality 0.832Customer Subscription Behaviour <------- Attitude 0.548Customer Subscription Behaviour <------- Service Quality 0.338

Table 9: Summary of Hypothesis TestingHypothesis Concluded StatusH1: Service Quality positively impacts buyer subscription behaviour. AcceptedH2: Service Quality positively impacts customer’s attitude towards Telecom Service Provider. AcceptedH3: Attitude positively impacts buyer subscription behaviour. AcceptedH4: Service quality positively impacts buyer subscription behaviour mediated by Customer’s attitude towards service provider. Accepted

African J. Basic & Appl. Sci., 9 (1): 07-18, 2017

15

Fig. 2: Structural Regression Model with Standardized Coefficients

Managerial Implications: Descriptive analysis clearly need to focus on ambience of service outlets, appearancecharts customer’s attitude towards the service of employees, their ability of problem solving, effectivedeterminants especially factors having propensity to communication with the customers and how do theyinfluence at service encounter stage. Therefore, managers behave with the customers.

African J. Basic & Appl. Sci., 9 (1): 07-18, 2017

16

According to the ‘Servicescape’ model, ambience is Customer subscription behavior but reduced to 39%of due importance [23, 55, 56, 57]. The service centers because of mediating effect of attitude. This study offersshould focus on making the physical infrastructure more new empirical support for the proposition that attitudeappealing in terms of lighting, color, sounds, smells, towards service provider play mediating role betweentemperature and availability of signage to guide perceived service quality and subscription behavior.customers through the service center. Moreover it’s As such, findings from this study aspires managers torecommended that all the service centers should look develop attitudinal strategies in order to increasesimilar, as it would reduce customer frustration of finding customer retention and acquisition in telecommunicationcounters. Customer waiting times can be effectively services.managed and reduced by increasing number of self-serving kiosks. REFERENCES

Employees are the very first contact point whichcustomers face and hence it matters that how they behave 1. Park, C. and Y. Kim, 2003. A framework of dynamicand act during service encounter. The facial expressions, CRM: linking marketing with information strategy,dressing style and overall personality matters a lot for the Bus Process Manag J., 9(5): 652-671. customers. One study shows that the front line employees 2. Rehman, A., R. Azam and A. Khattak, 2013.who were in professional dress were rated as good service Measuring validity of determinants of consumerquality as compared to the employees in casual dressing subscription behavior in telecom sector of pakistan.[31]. Managers should take the physical appearance of Jinnah Bus Rev, 1(2): 22-29.the employees as a serious issue in order to enhance the 3. Berry, L.L., A. Parasuraman and V.A. Zeithaml,service quality. Employee wellbeing programs can be a 1988. The service-quality puzzle. Bus Horiz.,viable option to increase their engagement and integrity 31(5): 35-43.with the organization. Special attention should be devoted 4. Bucak, T., 2014. The Effect of Service Quality onto employee training, development and career Customer Satisfaction?: A Research on Hotelprogression. Frequent service learning workshops, Businesses. Int. J. Educ Res., 2(1): 1-12. seminars, webinars and conferences should be organized 5. Kumar, J. and C.L. Yang, 2014. Service Quality andor employees should be encouraged to attend in order to Loyalty of International Students. Malaysian onlinemake their encounter with customers more pleasant. J. Serv. Qual., 2(3): 97-118. file:///C:/Users/zubair/Flexible working schedules can also reduce emotional AppData/Local/Mendeley.labor and pleasant service encounters can be the resulting 6. Parasuraman, A., V. Zeithaml, L. Berry, A.P, V.A.Z,outcome. L.L.B. SERVQUAL: A Multiple-Item Scale for

Telecom companies of UAE should start thinking Measuring Consumer Perceptions of Service Quality.about some innovative techniques to match the J. Retail., 1988; 64(1): 12-40.expectations. Companies should try to make 7. Kotler, P. and K. Keller, 2014. Marketingdifferentiation strategies so they would be able to serve Management Plus 2014 Mymarketinglab within a better way as compared to the competitors. Speed Pearson Etext -- Access Card Package. 14 ed.should be further enhanced by improving fiber optics P e a r s o n E d u c a t i o n ; 2 0 1 4 .network and by having more number of service centers https://books.google.com/books?id=86HKoAEACso that the traffic is controlled effectively and to reduce AAJ&pgis=1.the number of customers waiting on hold to get their 8. Grant, C., 1999. Theodore Levitt ’ s Marketingproblems solved. Myopia. 1999: 397-406. doi:10.1023/A.

Availability and provision of services at all-times and 9. Nilsson, L., M.D. Johnson and A. Gustafsson,through Omni-channels of access shall be ensured. An 2001. The impact of quality practices onincrease in quality, dependability, speed and flexibility of customer satisfaction and business results:services ensures sustainable competitive advantage and product versus service organizations, J. Qual.companies’ success in market. Manag., 6(1): 5-27.

CONCLUSION Customer-Perceived Service Quality in Fast-Food

Data analysis proves partial mediation. Service Satisfaction and Behavioral Intentions. Qual Managquality was having 92% positive direct influence on J., 15(2): 35-50.

th

10. Qin, H. and V.R. Prybutok, 2008. Determinants of

Restaurants and Their Relationship to Customer

African J. Basic & Appl. Sci., 9 (1): 07-18, 2017

17

11. Maglio, P.P., C.A. Kieliszewski and J.C. Spohrer, 23. Lovelock, C., P. Patterson and J. Wirtz, 2010. ServicesHandbook of Service Science. Springer Science & Marketing. Berkshire: Pearson Higher Education AU;Business Media; 2010. https://books.google.com/ https://books.google.com/books?id=NgbiBAAAQbooks?id=dndKmyps6fQC&pgis=1. BAJ&pgis=1.

12. Seth, N., S.G. Deshmukh and P. Vrat, 2005. Service 24. Yang, Z. and M. Jun, 2008. Consumer Perception ofquality models: a review, Int J. Qual Reliab Manag, E-Service Quality: From Internet Purchaser and Non-22(9): 913-949. purchaser Perspectives. J. Bus Strateg, 25(2): 59-84.

13. Palmer, A., 2014. Principles of Services Marketing: 25. Boulding, W., A. Kalra, R. Staelin and V. Zeithaml,Adrian Palmer: 9780077152345:Amazon.com: Books. 1993. A Dynamic Process Model of Service Quality:http://www.amazon.com/Principles-Services- From Expectations to Behavioral Intentions. J. MarkMarketing-Adrian-Palmer/dp/0077152344. Accessed Res., 30(1): 7. May 4, 2015. 26. Edvardsson, B., 2005 GURU’ S VIEW Service quality:

14. Lavidge, R.J. and G.A. Steiner, 2000. (Gary A. A beyond cognitive assessment. Manag Serv. Qual.,Model For Predictive Measurements of Advertising 15(2): 127-131.Effectiveness. Advert Soc. Rev., 1(1). 27. Siu, N.Y.M and J.T.H. Cheung, 2001. A measure of

15. Peter, J.P. and J.C. Olson, 2010. Consumer Behavior retail service quality. Mark Intell Plan., 19(2): 88-96.& Marketing Strategy. McGraw-Hill Higher 28. Parasuraman, A., 1994. Alternative scales forEduca t ion ; . https:/ /books.google.com/ measuring service quality: A comparativebooks?id=4D1VPgAACAAJ&pgis=1. assessment based on psychometric and diagnostic

16. Akbar, M.M. and N. Parvez, 2009. Impact of Service criteria, J. Retail., 70(3): 201-230. Quality, Trust and Customer Satisfaction on 29. Karatepe, O.M., U. Yavas and E. Babakus, 2005.Customers Loyalty. ABAC J.; http://its- Measuring service quality of banks: Scale3.au.edu/open_journal/index.php/abacjournal/articl development and validation, J. Retail Consum Serv.,e/view/526, 29(1). 12(5): 373-383.

17. Cronin, J.J., M.K. Brady and G.T.M. Hult, 2000. 30. Omotayo, O., 2011. Sales Promotion and ConsumerAssessing the effects of quality, value and customer Loyalty?: A Study of Nigerian Tecommunicationsatisfaction on consumer behavioral intentions in Industry, J. Compet, 4: 66-77.service environments, J. Retail., 76(2): 193-218. 31. Ramseook-Munhurrun, P., S.D. Lukea-Bhiwajee and

18. Mahapatra, S.C. and B.R. Shekhar, 2014. Developing P. Naidoo, Service Quality in the Public Service. 2010.a Measurement Model for Indian Airport Services http://papers.ssrn.com/abstract=1668833.Perceived by Air Passengers, Int J. Innov. Res. Dev., 32. Triandis, H.C., 1980. Values, attitudes and3(12). interpersonal behavior. In: Nebraska Symposium on

19. Kapiki, S. (Tania), 2012. Quality Management in Motivation, 27: 195-259.Tourism and Hospitality: an Exploratory Study 33. Ajzen, I. and M. Fishbein, 1980. UnderstandingamongTourism Stakeholders. Int J. Econ. Pract Attitudes and Predicting Social Behavior. New York;Theor., 2(2): 53-61. http://www.getcited.org/pub/103368990.

20. Angelova, B. and J. Zekiri, 2011. Measuring 34. Shiffman, L.G. and L.L. Kanuk, 2009. ConsumerCustomer Satisfaction with Service Quality Using Behaviour.American Customer Satisfaction Model (ACSI 35. Solomon, M., G. Bamossy, S. Askegaard andModel), Int J. Acad Res. Bus. Soc. Sci., 1(3): 27. M.K. Hogg, 2007. Consumer Behaviour: A European

21. Hernon, P. and D. Nitecki, 2001. Service Quality: Perspective.A Concept Not Fully Explored. Libr Trends, 36. Allport, G.W., 1935. Attitude. In: Murchison C, ed.49(4): 687-708. Handbook of Social Psychology. 1 ed. Worcester:

22. Boulding, W., A. Kalra, R .Staelin and V. Zeithaml, MA: Clark Univ. Press, pp: 798-844.1992. A dynamic process model of service quality: 37. Asiegbu, I.F., M. Powei Daubry and C. Iruka, 2012.From expectations to behavioral quality. J. Mark Consumer Attitude: Some Reflections on ItsRe s. ; 30. ht tp: / /search.ebscohost .com/ Concept, Trilogy, Relationship with Consumerlogin.aspx?direct=true&db=epref&AN=JMR.CJ.G. Behavior and Marketing Implications. Eur. J. Bus.BOULDING.DPMSQF&site=ehost-live. Manag., 4(13): 38-50.

st

African J. Basic & Appl. Sci., 9 (1): 07-18, 2017

18

38. Julian, C.C., 2008. Emotional Dissonance and 49. White, M.D. and E.G. Abels, 1994. Measuring QualityCustomer Service: An Exploratory Study. Serv Mark in Special Libraries: Lessons from Service Marketing.Q., 29(3): 1-23. Spec Libr, 86(1): 36-45.

39. Hagen M. Van, 2007. Association for European 50. Cesareo, L. and A. Pastore, 2014. Consumers’Transport and contributors 1. Transport, pp: 1-14. attitude and behavior towards online music piracy

40. Aaker, D. and K. Keller, 1990. Consumer Evaluations and subscription-based services, J Consum Mark.,of Brand Extensions, J Mark, 54(1). 31(6/7):515-525.http://search.ebscohost.com/login.aspx?direct=true 51. Sobel, M.E., 1982. Asymptotic confidence intervals&db=epref&AN=JM.ED.BG.AAKER.CEBE&site=e for indirect effects in structural equation models.host-live. Sociol Methodol, 13: 290-312.

41. Ariola, 2006. Al E. Principles and Methods of 52. Keppel, G., 1982. Design and Analysis: AResearch. Researcher’s Handbook. Englewood Cliffs. NJ:

42. Bentler, P.M. and C.P. Chou, 1987. Practical Prentice-Hall.Issues in Structural Modeling. Sociol Methods Res., 53. Saunders, M., P. Lewis and A. Thornhill, 2012.16(1): 78-117. Research Methods for Business Students.

43. Hair, J.F., W.C. Black, B.J. Babin and R.E. Anderson, 54. Khalil, O.E.M. and H.G.A. Ghanim, 2013. Perceived2014. Multivariate Data Analysis. 7 ed. EdinBurgh: Technical Service Quality and Informationth

Prentice-Hall. Satisfaction at the Ministry of Communications,44. Kline, R.B., 2015. Principles and Practice of Structural Kuwait. Inf. Resour Manag J., 26(2): 64-90.

Equation Modeling. Guilford Press; 2011. 55. Annafari, M.T. and E. Bohlin, 2014. Empiricalh t t p s : / / b o o k s . g o o g l e . c o m / b o o k s ? i d = - exploration of factors that determine multipleMDPILyu3DAC&pgis=1. Accessed May 3. mobile phone subscriptions, Int J. Mob. Commun.,

45. Byrne, B.M., 2015. Structural Equation Modeling 12(1): 12-28. With AMOS: Basic Concepts, Applications and 56. Bitner, M.J., 1992. Servicescapes: The Impact ofProgramming, Second Edition. Routledge; 2013. Physical Surroundings on Customers andhttps://books.google.com/books?id=0cHbAAAAQ Employees, J. Mark., 56(2). BAJ&pgis=1. Accessed May 3. 57. Baker, M. and S. Hart, 2007. The Marketing Book. 6th

46. Ajzen, I. and M. Fishbein, 1980. Understanding ed. Taylor & Francis.Attitudes and Predicting Social Behavior. New York.

47. Parasuraman, A., V. Zeithaml and L. Berry, 1985.A Conceptual Model of Service Quality and ItsImplications for Future Research. J. Mark.,49(4): 41-50.

48. Cronin, Jr. J.J., S.A. Taylor, J.J. Cronin andS.A. Taylor, 1994. SERVPERF Versus SERVQUAL:Reconciling Performance-Based and Perceptions-Minus-Expectations Measurement of Service Quality.Mark., 58: 125-131.